Clinical Outcomes Following Changes to a Nurse-Driven Diabetic Ketoacidosis Intravenous Insulin Protocol

Limited literature exists to support the safety and efficacy of a nurse-driven protocol for the management of diabetic ketoacidosis (DKA). Our institution’s nurse-driven protocol was updated in 2022 to improve outcomes and safety by adding standing rate calculations between lab assessments. We aimed to compare patient outcomes following changes made to the protocol at an academic medical center.

This was a retrospective, single-center, pre-post study that included adult patients (age 18 years or older) with a DKA diagnosis defined by the 2009 American Diabetes Association guidelines who had the nurse-driven DKA protocol ordered. The primary outcome was time from DKA diagnosis to resolution based on laboratory values. Secondary outcomes included frequency of hypoglycemic events, insulin infusion duration, patient disposition, and hospital length of stay. Comparisons between protocols were made using the Wilcoxon rank-sum test for continuous variables and the chi-square test for categorical variables.

We included 250 patients, according to sample size calculations, with 125 patients in the pre- and 125 patients in the post-implementation group. Patients who had a diagnosis of euglycemic DKA or hyperosmolar hyperglycemic state, pregnant at the time of presentation, transferred to or from an outside hospital, or presented to the pediatric emergency department were excluded. Preliminary results are available currently. There were significantly fewer patients with severe DKA in the pre-implementation group (pre- 61 [48.8%] versus post- 84 [67.2%], p=0.0032). There was no statistically significant difference in the time to DKA resolution between groups (pre- 15 hours [10.5-21.5] versus post- 16.5 hours [11-21], p=0.10). Hypoglycemic events were significantly higher in the pre-implementation group (25 [20%] versus 11 [8.8%], p=0.01). Insulin infusion duration and hospital length of stay were significantly longer in the post-implementation group.

Revision of the nurse-driven DKA protocol resulted in no statistical difference in time to DKA resolution, but was associated with a significant reduction in hypoglycemic events. Additional data analysis will be conducted to support these findings and will include an attempt to assess the time to resolution while controlling for differences in DKA severity.